Application of wavelet fractal features for the automated detection of epileptic seizure using electroencephalogram signals Online publication date: Wed, 02-Dec-2015
by Rahul Upadhyay; Swati Jharia; Prabin Kumar Padhy; Pavan Kumar Kankar
International Journal of Biomedical Engineering and Technology (IJBET), Vol. 19, No. 4, 2015
Abstract: In this paper, an attempt is made to find the appropriate wavelet function and wavelet-based fractal features for automated detection of epileptic seizure. Electroencephalogram (EEG) signals considered in this study include seizure and non-seizure EEG signals. Proposed study is occurred in four steps. In the first step, six frequency sub-bands of EEG signals (seizure and non-seizure) are computed using wavelet functions such as Haar, Biorthogonal (bior1.1 and bior2.2), Coiflets (coif1-coif3) and Daubechies (Db1-Db3). In the second step, wavelet thresholding is performed for undesirable noise suppression. Further, fractal dimensions are calculated from thresholded wavelet coefficients of four sub-bands as features in the third step. In the fourth step, the prepared feature vectors are fed to the artificial intelligence techniques for classifying seizure and non-seizure EEG signals. For classification three artificial intelligence techniques, i.e. least square-support vector machine, artificial neural network and random forest tree classifiers, are employed. Experimental result shows the effectiveness of the proposed methodology for epileptic seizure detection.
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Biomedical Engineering and Technology (IJBET):
Login with your Inderscience username and password:
Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.
If you still need assistance, please email subs@inderscience.com